Evidence linking cadmium and/or lead exposure to immunomodulatory effects in mammals based upon an adverse outcome pathways approach, and research perspectives

Chemosphere Evidence linking cadmium and/or lead exposure to immunomodulatory effects in mammals based upon an adverse outcome pathways approach, and research perspectives --Manuscript Draft-- Manuscript Number: CHEM142088R1 Article Type: Research paper Section/Category: Toxicology and Risk Assessment Keywords: AOPs; ecoimmunotoxicology; immunomodulation; oxidative stress; vertebrates; stress ecology Corresponding Author: Cloé HADJADJI Swiss Ornithological Institute Sempach, SWITZERLAND First Author: Cloé HADJADJI Order of Authors: Cloé HADJADJI Quentin Devalloir Colette Gaillard Nico van den Brink Renaud Scheifler Abstract: For decades, studies have shown how exposure to non-essential trace metals such as lead (Pb) and cadmium (Cd)impactwildlife. Ecoimmunotoxicology has emerged in the past two decades and focuses on the effects of pollutants on the immune system of free-ranging organisms. Adverse outcome pathways (AOPs) represent a conceptual approach to explore the mechanistic linkage between a molecular initiating event and adverse outcomes,at all biological levels of organisation. This paper proposes putative AOPs related to the effects of Cd, Pb, and the mixture Cd-Pb, on the immune system of mammals to address future questions in ecoimmunotoxicology. Molecular Initiating Events for both metals relate to entrance in cells through Ca2+ channels or bond to cell surfaces. Exposure to Cd, Pb and Cd-Pb share several similar Key Events (KEs), primarily an increase of oxidative stress (OS) in immune cells through production of reactive oxygen species. For both metals and the mixture, OS affects mitochondrial membranes, and induces apoptosis, ultimately decreasing immune cell number. Both metals affect innate immune system through nuclear factor kappa B (NF-κB) and mitogen-activated protein kinase (MAPK) inflammatory signalling pathways, leading to an upregulation of inflammatory markers and mediators. Adaptive immune system isaffected by the exposure to both metals though a decrease of CD4+/CD8+ ratioand MHCII, an inactivation of TH1 and TH2 response, and an inhibition of the humoral response mediated by various Ig. Mixture effects of Cd-Pb are less documented resulting in a more speculative AOP, butsynergic and antagonistic effects were identified. According to our AOPs, further research in ecoimmunotoxicology of metals inmammals should focus on KEs related to NF-κB/MAPK inflammatory signalling pathways, changes in CD4+/CD8+ ratio and MHCII complexes, and on AOs related to auto-immune disorders and on the effective increase of infection rate, particularly in case of exposure to metal mixtures. Suggested Reviewers: Nynke Kramer Wageningen UR [email protected] Willie Peijnenburg Universiteit Leiden [email protected] Christy Morrissey University of Saskatchewan [email protected] Manuel Ortiz Santaliestra IREC - Multidisciplinary Science for sustainable hunting and biodiversity conservation [email protected] Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation John Elliott Pacific Wildlife research center [email protected] Rikke Poulsen Aarhus Universitet [email protected] Opposed Reviewers: Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation 1 Cloe Hadjadji Swiss Ornithological Institute Seerose 1 CH-6204 Sempach And Laboratoire Chrono-environnement UMR 6249, CNRS/Université de Franche-Comté 16 route de Gray FR-25000 Besançon [email protected] Subject: Submission of a manuscript to Chemosphere Dear Editor, We are pleased to submit our manuscript titled “Adverse outcome pathways of trace metals (cadmium, lead) from molecular initiating events to immune dysfunctions in mammals”, which we would like to see published in Chemosphere. Among the approaches used in ecotoxicology to study the cascade (from molecular to population or community levels) of toxic effects of pollutants, adverse outcome pathways (AOPs) appear as an innovative and useful concept that could improve the mechanistic understanding of the impacts of contaminants on organisms. A few reviews addressed the single effects of these metals on the immune systems of mammals but, to our knowledge, none built proper adverse outcome pathways or reviewed the mixture effects of these metals on mammals’ immune response. In this paper, we built three different adverse outcome pathways about the single and combined effects of cadmium and lead that lead to immune dysfunctions in mammals. This manuscript gives new insights into how cadmium and lead exposure can modulate the immunocompetence of mammals using the AOPs framework. It allows us to identify clear research gaps in ecoimmunotoxicology. This work can serve current issues on pollution and pathogens interaction as well as guiding futures applied research and risk assessors. We therefore think that this paper would be of interest to the readership of Chemosphere. We propose below a list of potential reviewers for our article: - Nynke Kramer: [email protected] - Willie Peijnenburg: peijn[email protected]denuniv.nl - Christy Morrissey: [email protected] - Manuel Ortiz Santaliestra: [email protected] - John Elliott: [email protected] - Rikke Poulsen: [email protected] Thank you very much for considering our work. With our best regards, Cloe Hadjadji, on behalf of all co-authors. Besançon, 09/23/2024 Cover Letter 1 Evidence linking cadmium and/or lead exposure to immunomodulatory 1 effects in mammals based upon an adverse outcome pathways 2 approach, and research perspectives 3 C. Hadjadji1,2,*, Q. Devalloir2, C. Gaillard2, N. van den Brink3, R. Scheifler2 4 1Swiss ornithological institute, Seerose 1, CH-6204 Sempach, Switzerland 5 2Laboratoire Chrono-environnement, UMR 6249 CNRS / Université de Franche-Comté, 16 route de Gray, 6 25000 Besançon, France 7 3Division of Toxicology, Wageningen University, Box 8000 6700 EA, Wageningen, The Netherlands 8 9 *Corresponding author: Cloe Hadjadji, [email protected] 10 Abstract: For decades, studies have shown how exposure to non-essential trace metals such as lead (Pb) and 11 cadmium (Cd) largely impact global wildlife. Ecoimmunotoxicology has emerged in the past two decades and 12 focuses on the effects of pollutants on the immune system of free-ranging organisms. Adverse outcome 13 pathways (AOPs) represent a conceptual approach to explore the mechanistic linkage between a molecular 14 initiating event and adverse outcomes, potentially at all biological levels of organisation. The present paper 15 proposes putative AOPs related to the effects of Cd, Pb, and the mixture Cd-Pb, on the immune system of 16 mammals to address future questions in ecoimmunotoxicology. Molecular Initiating Events for both metals 17 relate to entrance in cells through Ca2+ channels or bond to cell surfaces. Exposure to Cd, Pb and Cd-Pb share 18 several similar Key Events (KEs), primarily an increase of oxidative stress (OS) in immune cells through 19 production of reactive oxygen species. For both metals and the mixture, OS affects mitochondrial membranes, 20 and induces apoptosis, ultimately decreasing immune cell number. Both metals affect innate immune system 21 through nuclear factor kappa B (NF-κB) and mitogen-activated protein kinase (MAPK) inflammatory signalling 22 pathways, leading to an upregulation of inflammatory markers and mediators. Adaptive immune system is 23 also affected by the exposure to both metals though a decrease of CD4+/CD8+ ratio, a decrease of MHCII, an 24 Title Page (with author names and complete affiliations) 2 inactivation of TH1 and TH2 response, and an inhibition of the humoral response mediated by various Ig. 25 Mixture effects of Cd-Pb are less documented resulting in a more speculative AOP, but potential synergic and 26 antagonistic effects were identified. According to our AOPs, further research in ecoimmunotoxicology of 27 metals in free-ranging mammals should focus on KEs related to NF-κB/MAPK inflammatory signalling 28 pathways, changes in CD4+/CD8+ ratio and MHCII complexes, and on AOs related to auto-immune disorders 29 and on the effective increase of infection rate, particularly in case of exposure to metal mixtures. 30 31 Highlights:  We built AOPs relevant to the effects of Cd and/or Pb on mammalian immune system  MIEs involve entrance in cells through Ca2+ channels or binding to cell surfaces  KEs involve oxidative stress, NF-κB signalling pathways, changes in CD4+/CD8+ ratio  AOs include immunodeficiency, auto-immunity, and potential increased infection rate  Gaps in current knowledge and detailed research perspectives are provided Highlights Manuscript Number: CHEM142088: Adverse outcome pathways of trace metals (cadmium, lead) from molecular initiating events to immune dysfunctions in mammals Besançon, 20/12/2024 Cloé Hadjadji Laboratoire Chrono-Environnement Université Bourgogne / Franche-Comté 16 route de Gray 25000 Besançon France & Swiss Ornithological Institute Seerose 1 CH-6204 Sempach Switzerland [email protected] Subject: Submission of the CHEM142088 revised manuscript Dear Editor, We are pleased to submit the revised version of our paper initially entitled “Adverse outcome pathways of trace metals (cadmium, lead) from molecular initiating events to immune dysfunctions in mammals”, which we would like to see published in Chemosphere. Please find hereafter the comments from Editors and Reviewers, and our responses. According to comments from the Reviewer 1, the title of the revised version would now be: “Evidence linking cadmium and/or lead exposure to immunomodulatory effects in mammals based upon an adverse outcome pathways approach, and research perspectives”. We thank the reviewers for their comments and suggestions, that we have followed carefully, as you will see below. In bold are the comments of the reviewers, in normal font our responses. We believe that reviewers’ comments indeed greatly improved our manuscript, and we hope that it is now acceptable for publication in Chemosphere. Thank you very much for considering our work. With our best regards, On behalf of co-authors of the revised version of the manuscript CHEM142088, Cloé Hadjadji. Response to Reviewers Reviewer #1: After reading through, I'm somewhat puzzled about the purpose of this paper. In broad strokes, the content is consistent with the objectives laid out in the introduction - namely to "propose putative AOPs" and "to address future questions in stress ecology" - although arguably, the authors "pose future questions" rather than addressing those questions. We understand the remark of the reviewer. Thus, we develop in the revised version a paragraph specific to our opinion about how the field of ecoimmunotoxicology of metals might be advanced. In the Introduction section, the objective of the present paper was stated as follows: “The present paper focusses on the effects of Cd and Pb on the immune system of mammals to propose putative AOPs to test more mechanistic hypotheses in ecoimmunotoxicology of metals and to address future questions in stress ecology (as other stressors than metals act on immunity).”. We acknowledge that perspectives were not developed enough in the previous section, and that the objective of “addressing future questions” was not fully fulfilled. Now that it has been done in the revised version, it seems that the sentence above clearly states the purpose of the paper, and we thus would like to keep it as it was. The authors conducted a literature review focused on immunomodulating effects of Cd and Pb exposure with the aim of developing hypothesized adverse outcome pathways. Notably, their approach ignores the first principle of AOP development - that AOPs are not chemical specific. Rather than focusing on general biological response to events like increased reactive oxygen species or inhibition of δ-ALAD activity, the authors outline highly overlapping, but nonetheless distinct AOPs for Cd, Pb, and a mixture of the two. Moreso than trying to develop chemically-agnostic generalized AOPs, it seems like the authors are using the AOP framework to organize an evidence map of the empirical results they found that potentially/plausibly link Cd or Pb exposure to immunodeficiency or immunostimulation. There is nothing fundamentally wrong with the latter, but I think the effort would be better portrayed as evidence mapping than AOP development. We also understand the reviewer’s remark. Our aim was not to develop AOPs as they indeed are, by definition, chemically-agnostic. Rather, we aimed at building what the reviewer calls an “evidence map” of the results published in the scientific literature about the mechanisms who link the exposure to Cd and Pb (or the mixture of both) to immunomodulatory effects in mammals. To do so, rather than a classical systematic literature review, we used the AOP approach as it allows to picture more mechanistically the relationships between exposure and effects. However, the field of ecoimmunotoxicology is relatively new, and thus we only found around 50 papers related to the issue. This prevented us to make true evidence mapping or to use the bibliometrics approach (as suggested by reviewer 2), as the latter is more appropriate to analyse a very large body of articles. The title of the revised version is now “Evidence linking cadmium and/or lead exposure to immunomodulatory effects in mammals based upon an adverse outcome pathways approach, and research perspectives”. This will prevent readers to think that we “developed” AOPs and understand more clearly that we used the AOP framework to address future research questions in the field of ecoimmunotoxicology. For the same reason, the titles of sections 3, 4 and 6 have been changed from “Development of an AOP for the effects of Cd on the immune system of mammals” towards “Evidence of the effects of Cd on the immune system of mammals based on an AOP framework”. If the goal were to develop AOPs, I would recommend that the authors populate information into the AOP-Wiki in accordance with OECD guidance and provide a full description of the key events and key event relationships (as available evidence allows) along with an overall evaluation of the level of support for the AOPs developed. Likewise, if the goal is AOP development, I would encourage the authors to focus more heavily on what is common to both Cd and Pb induced immunotoxicity and where the evidence is strongest across both stressors rather than on the margins or where the data are ambiguous. As said before, our aim was not really to develop AOP but rather to use an AOP approach to sum up what is known about the immunomodulatory effects of Cd and/or Pb in mammals, and then to draw research perspectives in this field. However, we acknowledge that having an AOP related to Cd exposure, and one related to Pb exposure, without comparing what is common or different between the 2 might also be frustrating to the reader. Thus, we added a new graph in the revised version tackling this comparison. We however would like to keep each of the AOP (and related paragraphs) as they allow to explain in more details the mechanisms related to the exposure of each of the 2 metals. In the latter half of the paper the authors describe some limitations and propose some next steps and broad questions that need to be addressed. However, much of this comes off as very broad generalizations and general calls that "we need to do this" but without any concrete or practical proposals for how to go about it. For example, lines 235-237 - the authors propose that "…every aspect of the potential impacts of toxic metals on the immune system….should be investigated in the wild…." How in practice, or in any practical sense would this be achieved? Such a statement is rather meaningless without any way to address it. Much of what follows details the complexities but offers no solutions. It's simple to say organisms and ecosystems are complex and challenge our understanding. It is far more difficult (but far more helpful) to offer practical strategies to cope with these challenges. The section concluded with "there is a great need to develop more immune-related methods in wild animals to achieve the same level of precision than in labs or humans" - without a strategy for how to do it, this is little more than words to fill a page. Its like saying, world peace is desirable - a nice sentiment, with little value if you can't provide a plan for how to achieve it. 2 mediated by toxic metals impact both innate and adaptive responses by unsettling the balanced mechanisms 27 of immune cell regulation (Tersago et al., 2004; Ebrahimi et al., 2020). Immunomodulation is also linked to a 28 myriad of external factors and stressors such as food shortage, life stage, or exposure to pathogens (Jackson 29 et al., 2009; Poulsen and Escher, 2012). Toxic effects also vary with species, sex, type (chronic or acute) and 30 route (ingestion, inhalation, dermal contact) of exposure (Gallucci et al., 2020). Immune cell dysfunction can 31 occur through ionic substitution of toxic metals due to similar affinity to receptors on organic compounds, 32 such as Cd2+ crossing Ca2+ channels (Choong et al., 2014). The main mechanism responsible for immune cell 33 damages is related to the toxicological impact of reactive oxygen species (ROS) including immune cell 34 membrane degradation (e.g., lipopolyperoxydation, Gera et al., 2015). Cadmium and Pb also have the ability 35 to upor downregulate inflammatory mediators and markers (cytokines, chemokines…), which can directly 36 modify the immune response (Ebrahimi et al., 2020; Devalloir et al., 2023). Cadmium and Pb can thus induce 37 differential effects on cytokine production through different signalling pathways (Thévenod and Lee, 2015). 38 For example, an increase of the pro-inflammatory response (TNF-α) has been shown in wood mice living near 39 a former smelter site contaminated by high Cd and Pb levels (Devalloir et al., 2023). 40 Since the emergence and the development of ecotoxicology in the 60’s, the discipline has evolved from a 41 relatively descriptive field to a much more mechanistic, hypothesis-based science. It aims at providing both a 42 better understanding of how pollutants can affect organisms at various spatial, temporal, and biological 43 organisation scales, and an improvement of risk assessment of chemicals released into the environment. 44 Among the approaches used in ecotoxicology to study the cascade (from molecular to population or 45 community levels) toxic effects of pollutants, adverse outcome pathways (AOPs) appeared as an innovative 46 and useful concept that could improve mechanistic approaches, here in ecoimmunotoxicology (Tryphonas, 47 2005; Baudiffier et al., 2024). Adverse outcome pathways represent a conceptual framework that highlights 48 existing knowledge linking a molecular initiating event (MIE) induced by a given toxic chemical to an adverse 49 outcome (AO) at multiple levels of biological organisation (Ankley et al., 2010). An AOP represents a 50 succession of key events (KEs) that can take various forms and depend on the information available. Molecular 51 initiating events are located at the molecular level and can be defined as “the initial points of chemical52 3 biological interaction within the organism” (Figure 1, Vinken et al., 2017). The relationship(s) between the 53 different events can be causal, mechanistic, inferential or correlation-based (Ankley et al., 2010). 54 Figure 1. 55 The present paper focusses on the effects of Cd and Pb on the immune system of mammals to propose 56 putative AOPs to test more mechanistic hypotheses in ecoimmunotoxicology of metals and to address future 57 questions in stress ecology (as other stressors than metals act on immunity) and ecotoxicology of free-ranging 58 mammals. Here, we focus on terrestrial mammals because reviews have recently been published on immune 59 toxicity of metals (or contaminants including metals) in fish (Lee et al., 2023, 2019), birds (Vallverdú-Coll et 60 al., 2019) and marine mammals (Desforges et al., 2016). All arsenic (As), Cd, Pb, and mercury (Hg) belong to 61 the 10 chemicals or groups of chemicals of major public health concern at a global scale (World Health 62 Organization, 2024). We however focus here on Cd and Pb only as they are more relevant than As or Hg to 63 terrestrial mammals in terms of exposure, and to keep the present article at a reasonable size. We based the 64 construction of our AOPs on reviewing the literature in the area, our own experience in ecoimmunotoxicology, 65 and different interactive databases related to AOPs. These databases provide open-source interfaces to 66 deposit AOPs, which facilitates the sharing of AOP knowledge and collaborative work (reviewed in Vinken et 67 al., 2017). 68 2. Materials and Methods 69 The articles presented in this paper have been selected according to the PRISMA statement (Moher et al., 70 2009). Literature search for study cases on terrestrial mammals was performed with Web of Sciences (June 71 2023) by using the following keywords, alone or in combination, full or truncated: immune system, immunity, 72 oxidative stress, AOPs/Adverse outcomes pathways, biomarkers, mammals (wild or captive), metals/heavy 73 metals, Pb/Lead, Cd/Cadmium, mixture effects. We also included other review articles about AOPs, 74 immunology and immunotoxicity in mammals. We ended up with 123 articles after the identification phase 75 and then selected the most pertinent articles based on their title and abstract (screening phase). After the 76 eligibility phase, we finally kept 51 articles, that were included in our synthesis. After reviewing the literature 77 and building pre-AOPs, we used the AOP-wiki interactive databases (https://aopwiki.org/, Vinken et al., 2017) 78 4 to expand our AOPs and link them to the existing knowledge. Each KE identified in the literature was 79 associated with a specific KE from AOP-wiki when it existed (Table 2). The Event ID of each KE was then added 80 in our AOPs (see Figures). General knowledge (i.e., introduction, information about TMs) and perspectives 81 were based on other bibliographic searches and on our experience. 82 Table 2. 83 3. Evidence of the effects of Cd on the immune system of mammals based on an AOP 84 framework 85 Cadmium in the environment comes from both natural and anthropogenic sources. Primary natural source 86 of Cd is related to geological processes, such as volcanic activity (Dœlsch et al., 2006). Its occurrence is natural 87 in all types of soils and minerals, e.g., sulfate, sulfide, carbonate… (Kabata-Pendias, 2010). Anthropogenic 88 sources of Cd include mining, smelting activity, transport, burial of nickel-Cd batteries, manufacturing, and 89 pesticides spreading (Hossein-Khannazer et al., 2020). Main routes of exposure to Cd in mammals are through 90 air, water, and food intake. Cadmium enters the body via the respiratory and the gastrointestinal tract (Mirkov 91 et al., 2021). It is then transported into the blood, bound to some proteins such as albumin (Mirkov et al., 92 2021) and brought to the organs where it accumulates. Overall, all the organs are targeted by Cd but its 93 accumulation is always higher in the liver and kidneys (Sarkar et al., 2013). Cells and tissues exposed to Cd 94 are mainly damaged by oxidative stress, even if reactive oxygen species (ROS) are not directly produced by 95 Cd. Reactive oxygen species are naturally produced during oxygen metabolism and are essential for cell 96 signalling and homeostasis. However, high amounts of ROS disrupt cell functioning, including immune cells, 97 and produce several adverse effects. Cadmium is considered as an immunotoxic inhibitor, influencing both 98 innate and adaptive immune responses, which can alter immune cells functions (Mirkov et al., 2021). Major 99 effects of Cd on the immune system reported in the literature are: (1) Cd-dependent apoptosis, (2) variations 100 in immune cells activation and differentiation, and (3) enhancement of inflammatory responses in immune 101 cells by upand down-regulation of cytokines production (Genchi et al., 2020; Wang et al., 2021). 102 We identified two main MIEs for Cd, which result in significant AOs on the immune system (Figure 2). The 103 first one is linked to the chemical properties of Cd as, under its ionic form Cd2+, it is known to interfere with 104 5 essential metal ions involved in biological functions of organisms (Thévenod and Lee, 2015). Thus, Cd2+ can 105 block Ca2+ channels because of their thiol group affinity. It induces a major disturbance of cellular homeostasis 106 and affects protein structure. The second MIE is the binding of Cd on the cell surface (Choong et al., 2014). 107 Both MIEs are activating a reaction cascade of several KEs linked to immune dysfunctions. 108 The first step of this cascade is the induction of ROS production. Increase in oxidative stress affects 109 mitochondrial membranes, which plays a major role in mitochondrial ROS formation, and leads to apoptosis 110 (Genchi et al., 2020; Wang et al., 2021). Reactive oxygen species increase is responsible for the activation of 111 the NF-κB pathway, which is in turn linked to the increase of the production of the pro-inflammatory cytokines 112 IL-1b, IL-6 and TNF-α (Wang et al., 2021) and of some chemokines (IL-8, MIP-2). Enhancement of both IL-6 113 and TNF-α in neutrophils (i.e., the most abundant type of granulocytes) leads to potential indirect effects of 114 Cd on haematopoiesis because these cytokines are regulators of this mechanism (Ulich et al., 1989). It can 115 indicate a hypersensitive reaction. Bioassays using macrophages from murine cells line RAW 264.7 exposed 116 to Cd concentrations showed an increase in the pro-inflammatory cytokine IL-1β associated with an inhibition 117 of IL-6 and IL-10 (Riemschneider et al., 2015). Because of this modulation, a potential hyper-reactivity of 118 macrophages can occur, which could lead to autoimmune disorder in mammals. In addition, Jung and Oh 119 (2019) reported an increase of TNF-α 6h after an exposure to Cd, which would indicate a response from 120 macrophages to Cd toxicity. However, 18h after the exposure, a decrease in TNF-α was associated with 121 apoptosis in RAW 264.7 (So et al., 2018). These results suggest that Cd immunotoxicity varies according to 122 exposure time, resulting in a hyperactivation of the immune system during short-term exposure while long123 term exposure results in immunodeficiency. 124 During the regulation of oxidative stress balance, the increase of glutathione (GSH) can cause a decrease 125 in CD4+/CD8+ ratio (CD: cluster of differentiation), with a decrease of CD4+ (helper T-cells) and an increase 126 of CD8+ (cytotoxic T-cells; Devalloir et al., submitted). Moreover, a decrease in CD4+ can be linked to a 127 diminution of IL-2 and IFN-γ production (Pathak and Khandelwal, 2009). A decrease in IL-2, the main factor 128 for proliferation of activated T-cells, has been observed in mouse splenic cells after Cd exposure (Pathak and 129 Khandelwal, 2009). These results show that Cd participates to the overall inactivation and immunodeficiency 130 of the adaptive immune response in mammals. In addition, IFN-γ can be both enhanced or inhibited by Cd 131 6 exposure (Krocova et al., 2000; Turley et al., 2019), depending on the dose and time-exposure. Turley et al. 132 (2019) reported an enhancement of IFN-γ during low-dose Cd exposure (32 ppm in drinking water for 10 days) 133 without significantly affecting T-cell polarization in rats. On the contrary, a decrease of IFN-γ has been 134 reported in T-cells when exposed to higher Cd concentrations (Krocova et al., 2000). 135 Reactive oxygen species production also promotes the expression of some pro-inflammatory cytokines 136 such as IL-1b, IL-6 and TNF-α and inhibits anti-inflammatory cytokines such as IL-10 in macrophages (Wang et 137 al., 2021). The overexpression of IL-1b, IL-6 and TNF-α as well as the inhibition of IL-10 can result in hyper138 inflammation, auto immune disorders and in rare cases, cancer (Thévenod and Lee, 2015). Overall, a 139 dysregulation in cytokine production can affect the activation of TH1, Th2 or TH17 (Devalloir, 2023). In 140 addition, the binding of Cd on the cell surface induces a decrease of the expression of the major 141 histocompatibility complex class 2 (MCHII) on B-cells, which could affect the presentation of pathogens to 142 antigen presenting cells and decrease the effectiveness of the immune system in response to pathogens. 143 Antibody production is also directly affected by Cd exposure. In the spleen, B-cell antibody responses differ 144 with dose, class of antigen (IgG, IgM), antigen type (T-cell independent, T-cell dependent) and exposure 145 duration (Fujimaki et al., 1982). A decrease in IgE production has been highlighted after Cd exposure in 146 humans (Wang et al., 2021). Cadmium exposure also induces the transformation of IgE antibodies into IgG 147 produced by B lymphocytes in humans (Marth et al., 2001), indicating an impairment of the cell activation 148 initiation during cytotoxic signals (Wang et al., 2021). 149 150 7 Figure 2. 151 4. Evidence of the effects of Pb on the immune system of mammals based on an AOP 152 framework 153 Lead is a naturally occurring but potentially harmful element, present at high concentrations in the 154 environment mainly because of anthropogenic sources. This element is naturally present in the earth crust at 155 very low concentration and small amount are released in the environment due to natural processes as rocks 156 weathering (Pattee and Pain, 2002). Regarding anthropogenic sources, Pb is used in paints, pesticides, 157 gasoline, batteries, and other emissions that can come from smelting activity and industry (Wani et al., 2015). 158 Main routes of exposure in mammals are through air and food intake, where it can be absorbed by respiratory 159 and digestive tracts (Balali-Mood et al., 2021). Accumulation in several organs generate multiple adverse 160 effects such as nervous system disorders (e.g., saturnism), kidney and liver alteration, anaemia, or learning 161 disorders in children (Ebrahimi et al., 2020). The oxidant-antioxidant balance system can be disturbed by Pb 162 exposure, which can lead to inflammatory responses in organs (Balali-Mood et al., 2021). Lead also is 163 considered as an immunotoxic inhibitor, influencing both innate and adaptive immune responses that can 164 alter all immune cells (Metryka et al., 2018). The major effects of Pb on the immune system reported in the 165 literature are (1) the enhancement of antiand pro-inflammatory responses, (2) depression of humoral 166 immunity, and (3) immune cell damage at high Pb concentration. 167 Lead exposure, either chronic or acute, can induce two potential MIEs, that result in severe AOs on the 168 functioning of the immune system. First, the inhibition of δ-aminolaevulinic acid dehydratase (δ-ALA-D) by Pb 169 can lead to a rapid oxidizing of δ-ALA. Quantification of δ-ALA-D has long been used as a biomarker of Pb 170 exposure (Rocha et al., 2012). However, its use as a biomarker of Pb exposure has to be cautious because 171 modulation of δ-ALA-D can occur after an exposure to other metals. When δ-ALA-D oxidation occurs, it 172 induces a production of free radicals including ROS, leading to oxidative stress (Lopes et al., 2016). Second, 173 Pb can bind to the cell surface to oxygen compounds due to its chemical properties. Both MIEs are activating 174 a reaction cascade of several KEs linked to immune responses dysfunctions (Figure 3). 175 8 Innate and adaptive immune responses are both impacted by a chronic or acute Pb exposure, by producing 176 ROS, modulating cytokines production, and TH1/TH2 polarisation. The induction of ROS production promotes 177 or suppresses the expression of a few cytokines. Iavicoli et al. (2006) reported an increase of IL-4 and a 178 decrease of IL-2 and IFN-γ production in mice exposed to high Pb level. In comparison, low-dose Pb exposure 179 induces a decrease in IL-4, and an increase of IL-2 and IFN-γ, meaning that immune response modulation 180 might be not linear (Iavicoli et al., 2006). These responses differently affect TH1 and TH2 polarisation and 181 humoral responses. High Pb levels could cause an impairment of cytotoxic signalling (TH1) while low-dose Pb 182 affects the coordination of helper T-cell response (TH2). 183 Studies also highlighted an increase of three pro-inflammatory cytokines, IL-1β, TNF-α and IL-6, after an 184 exposure to Pb (Fenga et al., 2017), which could promote auto-immune damage. However, antioxidant 185 activation associated with Pb exposure, in response to oxidative stress, modified pro-inflammatory responses 186 by decreasing IL-1β, TNF-α and IL-6. Furthermore, synergistic effect between Pb and lipopolysaccharides (i.e., 187 LPS; an inducer of inflammatory response in organisms) has been demonstrated in the literature (Cheng et 188 al., 2009). LPS-induced mortality under Pb exposure has been observed in animal studies (Dentener et al., 189 1989) and could be due to the tremendous release of TNF-α by different cells including both T and B 190 lymphocytes, macrophages, and monocytes. In human macrophages, in vitro stimulation by LPS showed a 191 surface TNF-α expression on over 50% of the cells after 24h. IL-8 gene expression is also enhanced by Pb 192 exposure when associated with p42/44 MAPK in human (Lin et al., 2015). Indeed, p42/44 MAPK has been 193 observed to activate IL-8 gene expression (Metryka et al., 2018). 194 Lead also causes a decline in humoral immunity by affecting both TH1 and TH2, which may increase host 195 vulnerability to bacterial and virus infections (Hemphill et al., 1971; Gainer, 1974; Luster et al., 1978). Lead 196 affects humoral immune response by reducing IgA and IgG production, which can induce inflammatory 197 diseases and cancers as long-term AOs (Anetor et al., 2008). Naïve CD4+ exposed to Pb will preferentially 198 differentiate in TH2 and lead to the inhibition of TH1 cell growth. This results in a decrease of IFN-γ and IgG2 199 with an increase of IL-4, IgE, and IgG4. Mice exposed to Pb and infected by Listeria monocytogenes showed 200 significant reduction in the CD4+ and CD8+ production (Dyatlov and Lawrence, 2002). 201 202 9 Figure 3. 203 5. Similarities and differences between Cd and Pb effects on the immune system of mammals 204 based on an AOP framework 205 Overall, Cd and Pb produce similar effects in vertebrates with outcomes resulting either 206 immunostimulation or immunodeficiency. Here, we propose a single AOP (Figure 4) presenting mechanisms 207 and pathways common to the exposure to both trace metals. To our knowledge, no experiments have been 208 conducted with the very same conditions to directly compare the immunotoxicological effect of Cd and Pb. 209 This means that we cannot be sure whether the differences in the immunomodulatory effects of the exposure 210 to Cd or Pb are actual differences in the underlying mechanisms or if this is related to a lack of research. 211 Concerning the innate system, common mechanisms include tissue inflammation related to the increased 212 production of pro-inflammatory cytokines within the NF-κB signalling pathway and a reduction of 213 phagocytosis activity. In both cases, cytokine concentrations are modulated either up or down, leading to an 214 increased inflammation or an impaired signalling of the immune response. Concerning the adaptive response, 215 both trace metals affect the expression of MHCII and produce a decrease of the CD4+/CD8+ ratio, leading to 216 an inhibition of immunoglobulin production and an impairment of Th1 and Th2 response. These various 217 common pathways all lead to three main adverse effects such as immunostimulation, potential auto-immune 218 disorder/damage, immunodeficiency leading to potential increase infection rate, and in some cases even 219 cancer. 220 Differences in immunomodulatory effects of the exposure to Cd or Pb rely mainly on one feature specific 221 to Pb (Inhibition of the δ aminolaevulinic acid dehydratase, ALA-D) and differences in cytokines mobilised 222 through exposure to Cd or Pb (see Figures 2 and 3, and text below). The inhibition of ALA-D increases oxidative 223 stress in cells (Rocha et al., 2012) but we did not find to what extend this contributes to the overall increase 224 of oxidative stress related to the exposure to Pb. Cytokines and chemokines are modulated by the exposure 225 to Cd or Pb but according to previous studies, most of them are different for each metal indicating that 226 different pathways of the immune response are involved. For example, the extra activation of NF-κB leads to 227 increased IL-8 and MIP-2 in the case of Cd exposure while an increase of IL-4 is reported in the case of Pb 228 10 exposure (Iavicoli et al., 2006; Wang et al., 2021). In the case of immunoglobulin production, studies showed 229 an inhibition of IgE and an inhibition of IgA and IgG in the case of Cd and Pb, respectively (Anetor et al., 2008; 230 Wang et al., 2021). 231 Figure 4. 232 6. Evidence of the effects of a Cd-Pb mixture on the immune system of mammals based on 233 an AOP framework 234 In the environment, organisms are generally exposed to a mixture of contaminants rather than to a single 235 trace metal. These multiple exposures can be detrimental to wild vertebrates even if contaminants are 236 present at low concentrations. This concept is called “cocktail effect” or “mixture toxicity” (Beyer et al., 2014). 237 Ecotoxicological studies have mainly been performed through a chemical-by-chemical approach of exposure, 238 while wildlife are exposed to “dynamic chemical mixtures” through their life (Morrissey et al., 2023). It exists 239 only a few papers that relate mixture effects of trace metal elements (Cd, Pb) on the immune system. Thus, 240 little is known about their additive, synergetic and/or antagonistic effects on the immune system of 241 vertebrates. Synergetic and antagonistic effects are the results of the competition among metals for organic 242 and inorganic ligands during accumulation and uptake (Altenburger et al., 2003; Belden et al., 2007; Balistrieri 243 and Mebane, 2014). It has been shown that Cd-Pb mixture can enhance the production of white blood cells 244 in mice, indicating the activation of the immune system in response to the exposure (Cobbina et al., 2015). In 245 addition, these authors found that the exposure to a mixture result in a global antagonistic effect, in 246 comparison to their individual toxic effect. One hypothesis could be that Cd and Pb both produce thiols such 247 as GSH, enabling a protection of the cell from their toxicity. However, other antioxidant markers (superoxide 248 dismutase, glutathione peroxidase) were lower when mice were exposed to a Cd and Pb mixture, potentially 249 creating higher oxidative stress in immune cells. Another study in male rat showed that the exposure to a 250 mixture of eight trace metals, including Cd and Pb, induces a suppression of both humoral and cell-mediated 251 immune response (Jadhav et al., 2007). Fortier et al. (2008) depicted a non-significant effect of the Cd-Pb 252 mixture in human leucocytes (viability and apoptosis proportion). These few studies provide different 253 conclusions, suggesting that a variety of pathways could be involved in the response to Cd-Pb mixture. 254 11 However, we still lack information about how cytokines signalling, TH1/TH2 polarization of B-cells and antibody 255 production could be modulated by this mixture. Here, we propose a potential AOP for Cd-Pb mixture on the 256 immune system of mammals with potential additive effects or synergetic and antagonistic interactions (Figure 257 5), based on the few studies presented earlier as well as on hypotheses and deductions from our single258 chemical AOPs (Figures 2, 3 and 4). 259 260 18 consideration of the general chemical complexity (i.e., mixture effects, chemical properties, metabolites) of 422 pollutants. Each compound has its own mode(s) of action and pathways that lead to one or several toxic 423 effects at different levels of biological organisation. Identifying molecular initiating event(s) and key event(s) 424 that lead to adverse outcome(s) participates to understand how complex are the interactions between 425 pollutants, here Cd and Pb, in organisms. Second, once AOPs are built and the main pathways identified, it is 426 possible to make it speciesor life-stages specific (i.e., considering organismal complexity; Morrissey et al., 427 2023). This will guide the formulation of hypothesis and the choice of biomarkers. This (new) framework 428 participates in answering new questions and guide researchers into new directions in ecotoxicology, stress 429 ecology and ecoimmunotoxicology. Building AOPs before starting an experiment appears as something that 430 should become more common in these field of research to better apprehend mechanisms underlying toxic 431 effects and identify the research gaps (Baudiffier et al., 2024). In addition, developing the knowledge of the 432 database presented by Vinken et al., (2017) would encourage data sharing and literature findings to better 433 direct further research. To date, the most important limitations in making AOPs about a specific subject is the 434 lack of literature that can restrict the whole process. 435 436 8. Conclusion 437 Adverse outcome pathways are an efficient method to identify the mechanisms underlying the effects of 438 pollutants at various levels of biological organisation and thus, to identify research gaps in ecotoxicology and 439 risk assessment. Here, we identified potential MIEs and KEs that lead to AOs related to immune disorders in 440 mammals in response to Cd and/or Pb exposure. Cadmium and Pb may induce immuno-stimulation and 441 immunodeficiency through different pathways of the immune system, impacting both innate and adaptive 442 immune responses. Our findings could guide future research in ecoimmunotoxicology of metals. There are 443 still large open questions about how organisms, at different levels of organisation, react to pathogens and 444 infections in a context where exposure to pollutants affects immunity. Mixture effects of trace metal are also 445 still largely understudied. It is now of major importance to understand how these pollutants interact with 446 wildlife to improve conservation policies. We identified potential synergetic and antagonistic interactions 447 19 produced by the co-exposure to Cd and Pb, which also brings questions about the potential adaptation of 448 wild populations to those mixtures, in comparison to single-pollutant exposure. 449 Acknowledgements 450 The authors thank Michaël Coeurdassier, Clémentine Fritsch, and Francis Raoul, for fruitful discussion about 451 the present article. 452 Author contributions 453 Conceptualization: CH, RS; Data curation: CH; Formal analysis: CH; Funding acquisition: RS; Investigation: CH; 454 Methodology: CH; Project administration: RS; Supervision: QD, NvdB, RS; Validation: CH, QD, CG, NvdB, RS; 455 Visualization: CH; Writing - original draft: CH, RS; Writing - review & editing: CH, QD, CG, NvdB, RS. 456 Declaration of competing interest 457 The authors declare no competing interests. 458 Funding information 459 This study is funded by the CHRONIC project (No. 956009) ‘Chronic exposure scenarios driving environmental 460 risks of Chemicals’, an Innovative Training Network (ITN) funded by the European Union's Horizon 2020 461 research and innovation program under the Marie Skłodowska-Curie Actions (MSCA). 462 463 20 References 464 Abolins, S., King, E.C., Lazarou, L., Weldon, L., Hughes, L., Drescher, P., Raynes, J.G., Hafalla, J.C.R., Viney, M.E., 465 Riley, E.M., 2017. The comparative immunology of wild and laboratory mice, Mus musculus 466 domesticus. Nat. Commun. 8, 14811. https://doi.org/10.1038/ncomms14811 467 Abolins, S., Lazarou, L., Weldon, L., Hughes, L., King, E.C., Drescher, P., Pocock, M.J.O., Hafalla, J.C.R., Riley, 468 E.M., Viney, M., 2018. The ecology of immune state in a wild mammal, Mus musculus domesticus. 469 PLOS Biol. 16, e2003538. https://doi.org/10.1371/journal.pbio.2003538 470 Abolins, S.R., Pocock, M.J.O., Hafalla, J.C.R., Riley, E.M., Viney, M.E., 2011. Measures of immune function of 471 wild mice, Mus musculus. Mol. Ecol. 20, 881–892. https://doi.org/10.1111/j.1365472 294X.2010.04910.x 473 Altenburger, R., Nendza, M., Schüürmann, G., 2003. Mixture toxicity and its modeling by quantitative 474 structure-activity relationships. Environ. Toxicol. Chem. 22, 1900–1915. https://doi.org/10.1897/01475 386 476 Anetor, J.I., Anetor, G.O., Iyanda, A.A., Adeniyi, F., 2008. Environmental chemicals and human neurotoxicity: 477 magnitude, prognosis and markers. Afr. J. Biomed. Res. 11. https://doi.org/10.4314/ajbr.v11i1.50675 478 Ankley, G.T., Bennett, R.S., Erickson, R.J., Hoff, D.J., Hornung, M.W., Johnson, R.D., Mount, D.R., Nichols, J.W., 479 Russom, C.L., Schmieder, P.K., Serrrano, J.A., Tietge, J.E., Villeneuve, D.L., 2010. Adverse outcome 480 pathways: A conceptual framework to support ecotoxicology research and risk assessment. Environ. 481 Toxicol. Chem. 29, 730–741. https://doi.org/10.1002/etc.34 482 Aras, N.K., Ataman, O.Y., 2007. Trace Element Analysis of Food and Diet. Royal Society of Chemistry. 483 Balali-Mood, M., Naseri, K., Tahergorabi, Z., Khazdair, M.R., Sadeghi, M., 2021. Toxic Mechanisms of Five 484 Heavy Metals: Mercury, Lead, Chromium, Cadmium, and Arsenic. Front. Pharmacol. 12, 643972. 485 https://doi.org/10.3389/fphar.2021.643972 486 Balistrieri, L.S., Mebane, C.A., 2014. Predicting the toxicity of metal mixtures. Sci. Total Environ. 466–467, 487 788–799. https://doi.org/10.1016/j.scitotenv.2013.07.034 488 Baudiffier, D., Audouze, K., Armant, O., Frelon, S., Charles, S., Beaudouin, R., Cosio, C., Payrastre, L., Siaussat, 489 D., Burgeot, T., Mauffret, A., Degli Esposti, D., Mougin, C., Delaunay, D., Coumoul, X., 2024. Editorial 490 trend: adverse outcome pathway (AOP) and computational strategy — towards new perspectives in 491 ecotoxicology. Environ. Sci. Pollut. Res. 31, 6587–6596. https://doi.org/10.1007/s11356-023-30647492 w 493 Belden, J.B., Gilliom, R.J., Lydy, M.J., 2007. How well can we predict the toxicity of pesticide mixtures to 494 aquatic life? Integr. Environ. Assess. Manag. 3, 364–372. https://doi.org/10.1002/ieam.5630030307 495 Bergner, L.M., Orton, R.J., Da Silva Filipe, A., Shaw, A.E., Becker, D.J., Tello, C., Biek, R., Streicker, D.G., 2019. 496 Using noninvasive metagenomics to characterize viral communities from wildlife. Mol. Ecol. Resour. 497 19, 128–143. https://doi.org/10.1111/1755-0998.12946 498 Beyer, J., Petersen, K., Song, Y., Ruus, A., Grung, M., Bakke, T., Tollefsen, K.E., 2014. Environmental risk 499 assessment of combined effects in aquatic ecotoxicology: A discussion paper. Mar. Environ. Res., 500 Pollutant Responses in Marine Organisms” (PRIMO17) 96, 81–91. 501 https://doi.org/10.1016/j.marenvres.2013.10.008 502 Birnie-Gauvin, K., Peiman, K.S., Raubenheimer, D., Cooke, S.J., 2017. Nutritional physiology and ecology of 503 wildlife in a changing world. Conserv. Physiol. 5. https://doi.org/10.1093/conphys/cox030 504 Ceballos, G., Ehrlich, P.R., Dirzo, R., 2017. Biological annihilation via the ongoing sixth mass extinction signaled 505 by vertebrate population losses and declines. Proc. Natl. Acad. Sci. 114. 506 https://doi.org/10.1073/pnas.1704949114 507 Choong, G., Liu, Y., Templeton, D.M., 2014. Interplay of calcium and cadmium in mediating cadmium toxicity. 508 Chem. Biol. Interact. 211, 54–65. https://doi.org/10.1016/j.cbi.2014.01.007 509 Cobbina, S.J., Chen, Y., Zhou, Z., Wu, Xueshan, Zhao, T., Zhang, Z., Feng, W., Wang, W., Li, Q., Wu, Xiangyang, 510 Yang, L., 2015. Toxicity assessment due to sub-chronic exposure to individual and mixtures of four 511 toxic heavy metals. J. Hazard. Mater. 294, 109–120. https://doi.org/10.1016/j.jhazmat.2015.03.057 512 Costantini, D., 2022. A meta-analysis of impacts of immune response and infection on oxidative status in 513 vertebrates. Conserv. Physiol. 10, coac018. https://doi.org/10.1093/conphys/coac018 514 21 Dentener, M.A., Greve, J.W., Maessen, J.G., Buurman, W.A., 1989. Role of Tumour Necrosis Factor in the 515 Enhanced Sensitivity of Mice to Endotoxin After Exposure to Lead. Immunopharmacol. 516 Immunotoxicol. 11, 321–334. https://doi.org/10.3109/08923978909005373 517 Desforges, J.-P.W., Sonne, C., Levin, M., Siebert, U., De Guise, S., Dietz, R., 2016. Immunotoxic effects of 518 environmental pollutants in marine mammals. Environ. Int. 86, 126–139. 519 https://doi.org/10.1016/j.envint.2015.10.007 520 Devalloir, Q., 2023. Effects of multi-stressors (pollution, nutritional quality) on the immunocompetence of 521 wood mice. 522 Devalloir, Q., Fritsch, C., Alchammas, Y., Raoul, F., Driget, V., Amiot, C., Ozaki, S., van den Brink, N., Scheifler, 523 R., 2023. Environmental pollution and nutritional quality modulate immune response of the wood 524 mouse (Apodemus sylvaticus) through hormonal disturbances. Environ. Pollut. 337, 122100. 525 https://doi.org/10.1016/j.envpol.2023.122100 526 Devalloir, Q., Fritsch, C., Raoul, F., Ozaki, S., Alchammas, Y., n.d. The effect of toxic metals vs micronutrients 527 on mammal immunity. 528 Diagne, C., Ribas, A., Charbonnel, N., Dalecky, A., Tatard, C., Gauthier, P., Haukisalmi, V., Fossati-Gaschignard, 529 O., Bâ, K., Kane, M., Niang, Y., Diallo, M., Sow, A., Piry, S., Sembène, M., Brouat, C., 2016. Parasites 530 and invasions: changes in gastrointestinal helminth assemblages in invasive and native rodents in 531 Senegal. Int. J. Parasitol. 46, 857–869. https://doi.org/10.1016/j.ijpara.2016.07.007 532 Dœlsch, E., Saint Macary, H., Van de Kerchove, V., 2006. Sources of very high heavy metal content in soils of 533 volcanic island (La Réunion). J. Geochem. Explor. 88, 194–197. 534 https://doi.org/10.1016/j.gexplo.2005.08.037 535 Dyatlov, V.A., Lawrence, D.A., 2002. Neonatal Lead Exposure Potentiates Sickness Behavior Induced by Listeria 536 monocytogenes Infection of Mice. Brain. Behav. Immun. 16, 477–492. 537 https://doi.org/10.1006/brbi.2001.0641 538 Ebrahimi, M., Khalili, Neda, Razi, S., Keshavarz-Fathi, M., Khalili, Nastaran, Rezaei, N., 2020. Effects of lead and 539 cadmium on the immune system and cancer progression. J. Environ. Health Sci. Eng. 18, 335–343. 540 https://doi.org/10.1007/s40201-020-00455-2 541 Eskeland, S., Bø-Granquist, E.G., Stuen, S., Lybeck, K., Wilhelmsson, P., Lindgren, P.-E., Makvandi-Nejad, S., 542 2023. Temporal patterns of gene expression in response to inoculation with a virulent Anaplasma 543 phagocytophilum strain in sheep. Sci. Rep. 13, 20399. https://doi.org/10.1038/s41598-023-47801-6 544 Fenga, C., Gangemi, S., Di Salvatore, V., Falzone, L., Libra, M., 2017. Immunological effects of occupational 545 exposure to lead (Review). Mol. Med. Rep. 15, 3355–3360. https://doi.org/10.3892/mmr.2017.6381 546 Flies, A.S., Wild Comparative Immunology Consortium, 2020. Rewilding immunology. Science 369, 37–38. 547 https://doi.org/10.1126/science.abb8664 548 Fortier, M., Omara, F., Bernier, J., Brousseau, P., Fournier, M., 2008. Effects of Physiological Concentrations of 549 Heavy Metals Both Individually and in Mixtures on the Viability and Function of Peripheral Blood 550 Human Leukocytes In Vitro. J. Toxicol. Environ. Health A 71, 1327–1337. 551 https://doi.org/10.1080/15287390802240918 552 Fujimaki, H., Murakami, M., Kubota, K., 1982. In vitro evaluation of cadmium-induced augmentation of the 553 antibody response. Toxicol. Appl. Pharmacol. 62, 288–293. https://doi.org/10.1016/0041554 008X(82)90127-2 555 Gainer, J.H., 1974. Lead Aggravates Viral Disease and Represses the Antiviral Activity of Interferon Inducers. 556 Environ. Health Perspect. 7, 113–119. https://doi.org/10.1289/ehp.747113 557 Gallucci, R., Luckett-Chastain, L., Yucesoy, B., 2020. Immunotoxicology, in: An Introduction to Interdisciplinary 558 Toxicology. Elsevier, pp. 233–244. https://doi.org/10.1016/B978-0-12-813602-7.00018-1 559 García-Mendoza, D., Van Den Berg, H.J.H.J., Van Den Brink, N.W., 2021. Environmental exposure to cadmium 560 reduces the primary antibody-mediated response of wood mice (Apodemus sylvaticus) from 561 differentially polluted locations in the Netherlands. Environ. Pollut. 289, 117909. 562 https://doi.org/10.1016/j.envpol.2021.117909 563 Genchi, G., Sinicropi, M.S., Lauria, G., Carocci, A., Catalano, A., 2020. The Effects of Cadmium Toxicity. Int. J. 564 Environ. Res. Public. Health 17, 3782. https://doi.org/10.3390/ijerph17113782 565 Gera, R., Singh, V., Sharma, A.K., Ghosh, D., 2015. Chapter Four - Immunomodulation by Environmental 566 Chemicals: Insights into Mammalian Immune Responses to Arsenic, Cadmium, and Lead, in: Fishbein, 567 22 J.C., Heilman, J.M. (Eds.), Advances in Molecular Toxicology. Elsevier, pp. 109–159. 568 https://doi.org/10.1016/B978-0-12-802229-0.00004-9 569 Hemphill, F.E., Kaeberle, M.L., Buck, W.B., 1971. Lead Suppression of Mouse Resistance to Salmonella 570 typhimurium. Science 172, 1031–1032. https://doi.org/10.1126/science.172.3987.1031 571 Hossein-Khannazer, N., Azizi, G., Eslami, S., Alhassan Mohammed, H., Fayyaz, F., Hosseinzadeh, R., Usman, 572 A.B., Kamali, A.N., Mohammadi, H., Jadidi-Niaragh, F., Dehghanifard, E., Noorisepehr, M., 2020. The 573 effects of cadmium exposure in the induction of inflammation. Immunopharmacol. Immunotoxicol. 574 42, 1–8. https://doi.org/10.1080/08923973.2019.1697284 575 Hotaling, S., Kelley, J.L., Frandsen, P.B., 2021. Toward a genome sequence for every animal: Where are we 576 now? Proc. Natl. Acad. Sci. 118, e2109019118. https://doi.org/10.1073/pnas.2109019118 577 Iavicoli, I., Carelli, G., Stanek, E.J., Castellino, N., Calabrese, E.J., 2006. Below background levels of blood lead 578 impact cytokine levels in male and female mice. Toxicol. Appl. Pharmacol. 210, 94–99. 579 https://doi.org/10.1016/j.taap.2005.09.016 580 Jackson, J.A., Friberg, I.M., Bolch, L., Lowe, A., Ralli, C., Harris, P.D., Behnke, J.M., Bradley, J.E., 2009. 581 Immunomodulatory parasites and toll-like receptor-mediated tumour necrosis factor alpha 582 responsiveness in wild mammals. BMC Biol. 7, 16. https://doi.org/10.1186/1741-7007-7-16 583 Jadhav, S.H., Sarkar, S.N., Ram, G.C., Tripathi, H.C., 2007. Immunosuppressive Effect of Subchronic Exposure 584 to a Mixture of Eight Heavy Metals, Found as Groundwater Contaminants in Different Areas of India, 585 Through Drinking Water in Male Rats. Arch. Environ. Contam. Toxicol. 53, 450–458. 586 https://doi.org/10.1007/s00244-006-0177-1 587 Kabata-Pendias, A., 2010. Trace Elements in Soils and Plants, 4th ed. CRC Press, Boca Raton. 588 https://doi.org/10.1201/b10158 589 Krocova, Z., Macela, A., Kroca, M., Hernychova, L., 2000. The immunomodulatory effect(s) of lead and 590 cadmium on the cells of immune system in vitro. Toxicol. In Vitro 14, 33–40. 591 https://doi.org/10.1016/S0887-2333(99)00089-2 592 Lanno, R., Wells, J., Conder, J., Bradham, K., Basta, N., 2004. The bioavailability of chemicals in soil for 593 earthworms. Ecotoxicol. Environ. Saf. 57, 39–47. https://doi.org/10.1016/j.ecoenv.2003.08.014 594 Lee, J.-W., Choi, H., Hwang, U.-K., Kang, J.-C., Kang, Y.J., Kim, K.I., Kim, J.-H., 2019. Toxic effects of lead exposure 595 on bioaccumulation, oxidative stress, neurotoxicity, and immune responses in fish: A review. Environ. 596 Toxicol. Pharmacol. 68, 101–108. https://doi.org/10.1016/j.etap.2019.03.010 597 Lee, J.-W., Jo, A.-H., Lee, D.-C., Choi, C.Y., Kang, J.-C., Kim, J.-H., 2023. Review of cadmium toxicity effects on 598 fish: Oxidative stress and immune responses. Environ. Res. 236, 116600. 599 https://doi.org/10.1016/j.envres.2023.116600 600 Lin, Y.-C., Wei, P.-L., Tsai, Y.-T., Wong, J.-H., Chang, C.-M., Wang, J.-Y., Hou, M.-F., Lee, Y.-C., Chuang, H.-Y., Chang, 601 W.-C., 2015. Pb2+ induced IL-8 gene expression by extracellular signal-regulated kinases and the 602 transcription factor, activator protein 1, in human gastric carcinoma cells. Environ. Toxicol. 30, 315– 603 322. https://doi.org/10.1002/tox.21909 604 Liu, T., Zhang, L., Joo, D., Sun, S.-C., 2017. NF-κB signaling in inflammation. Signal Transduct. Target. Ther. 2, 605 1–9. https://doi.org/10.1038/sigtrans.2017.23 606 Lopes, A.C.B.A., Peixe, T.S., Mesas, A.E., Paoliello, M.M.B., 2016. Lead Exposure and Oxidative Stress: A 607 Systematic Review, in: de Voogt, P. (Ed.), Reviews of Environmental Contamination and Toxicology 608 Volume 236, Reviews of Environmental Contamination and Toxicology. Springer International 609 Publishing, Cham, pp. 193–238. https://doi.org/10.1007/978-3-319-20013-2_3 610 Luster, M.I., Faith, R.E., Kimmel, C.A., 1978. Depression of humoral immunity in rats following chronic 611 developmental lead exposure. J. Environ. Pathol. Toxicol. 1, 397–402. 612 Luster, M.I., Gerberick, G.F., 2010. Immunotoxicology Testing: Past and Future, in: Dietert, R.R. (Ed.), 613 Immunotoxicity Testing: Methods and Protocols, Methods in Molecular BiologyTM. Humana Press, 614 Totowa, NJ, pp. 3–13. https://doi.org/10.1007/978-1-60761-401-2_1 615 Marth, E., Jelovcan, S., Kleinhappl, B., Gutschi, A., Baryh, S., 2001. The Effect of Heavy Metals on the Immune 616 System at Low Concentrations. Int J Occup Med Env. Health 14, 375–386. 617 McKee, A.S., Fontenot, A.P., 2016. Interplay of innate and adaptive immunity in metal-induced 618 hypersensitivity. Curr. Opin. Immunol., Allergy and hypersensitivity * Host pathogens 42, 25–30. 619 https://doi.org/10.1016/j.coi.2016.05.001 620 23 Meier-Soelch, J., Mayr-Buro, C., Juli, J., Leib, L., Linne, U., Dreute, J., Papantonis, A., Schmitz, M.L., Kracht, M., 621 2021. Monitoring the Levels of Cellular NF-κB Activation States. Cancers 13, 5351. 622 https://doi.org/10.3390/cancers13215351 623 Metryka, E., Chibowska, K., Gutowska, I., Falkowska, A., Kupnicka, P., Barczak, K., Chlubek, D., Baranowska624 Bosiacka, I., 2018. Lead (Pb) Exposure Enhances Expression of Factors Associated with Inflammation. 625 Int. J. Mol. Sci. 19, 1813. https://doi.org/10.3390/ijms19061813 626 Mirkov, I., Popov Aleksandrov, A., Ninkov, M., Tucovic, D., Kulas, J., Zeljkovic, M., Popovic, D., Kataranovski, M., 627 2021. Immunotoxicology of cadmium: Cells of the immune system as targets and effectors of 628 cadmium toxicity. Food Chem. Toxicol. 149, 112026. https://doi.org/10.1016/j.fct.2021.112026 629 Moher, D., Liberati, A., Tetzlaff, J., Altman, D.G., 2009. Preferred Reporting Items for Systematic Reviews and 630 Meta-Analyses: The PRISMA Statement. Ann. Intern. Med. 151, 264–269. 631 https://doi.org/10.7326/0003-4819-151-4-200908180-00135 632 Morrissey, C., Fritsch, C., Fremlin, K., Adams, W., Borgå, K., Brinkmann, M., Eulaers, I., Gobas, F., Moore, D.R.J., 633 Van Den Brink, N., Wickwire, T., 2023. Advancing exposure assessment approaches to improve wildlife 634 risk assessment. Integr. Environ. Assess. Manag. ieam.4743. https://doi.org/10.1002/ieam.4743 635 Naidu, R., Bolan, N.S., Megharaj, M., Juhasz, A.L., Gupta, S.K., Clothier, B.E., Schulin, R., 2008. Chapter 1 636 Chemical bioavailability in terrestrial environments, in: Developments in Soil Science. Elsevier, pp. 1– 637 6. https://doi.org/10.1016/S0166-2481(07)32001-1 638 Pathak, N., Khandelwal, S., 2009. Immunomodulatory role of piperine in cadmium induced thymic atrophy 639 and splenomegaly in mice. Environ. Toxicol. Pharmacol. 28, 52–60. 640 https://doi.org/10.1016/j.etap.2009.02.003 641 Pattee, O.H., Pain, D.J., 2002. Lead in the Environment, in: Handbook of Ecotoxicology. CRC Press. 642 Poulsen, A., Escher, B., 2012. Chemically induced immunosuppression and disease susceptibility in marine 643 wildlife: A literature review. Entox Natl. Res. Cent. Environ. Toxicol. Univ. Qld. 110. 644 Powolny, T., Scheifler, R., Raoul, F., Coeurdassier, M., Fritsch, C., 2023. Effects of chronic exposure to toxic 645 metals on haematological parameters in free-ranging small mammals. Environ. Pollut. 317, 120675. 646 https://doi.org/10.1016/j.envpol.2022.120675 647 Punt, J., 2019. Kuby immunology. New York : Macmillan Education. 648 Riemschneider, S., Herzberg, M., Lehmann, J., 2015. Subtoxic Doses of Cadmium Modulate Inflammatory 649 Properties of Murine RAW 264.7 Macrophages. BioMed Res. Int. 2015, e295303. 650 https://doi.org/10.1155/2015/295303 651 Rocha, J.B.T., Saraiva, R.A., Garcia, S.C., Gravina, F.S., Nogueira, C.W., 2012. Aminolevulinate dehydratase (δ652 ALA-D) as marker protein of intoxication with metals and other pro-oxidant situations. Toxicol. Res. 1, 653 85. https://doi.org/10.1039/c2tx20014g 654 Sarkar, A., Ravindran, G., Krishnamurthy, V., 2013. A brief review on the effect of cadmium toxicity: from 655 cellular to organ level. Int J Biotechnol Res 3, 17–36. 656 So, K.-Y., Lee, B.-H., Oh, S.-H., 2018. The critical role of autophagy in cadmium-induced immunosuppression 657 regulated by endoplasmic reticulum stress-mediated calpain activation in RAW264.7 mouse 658 monocytes. Toxicology 393, 15–25. https://doi.org/10.1016/j.tox.2017.10.016 659 Tanabe, S., O’Brien, J., Tollefsen, K.E., Kim, Y., Chauhan, V., Yauk, C., Huliganga, E., Rudel, R.A., Kay, J.E., Helm, 660 J.S., Beaton, D., Filipovska, J., Sovadinova, I., Garcia-Reyero, N., Mally, A., Poulsen, S.S., Delrue, N., 661 Fritsche, E., Luettich, K., La Rocca, C., Yepiskoposyan, H., Klose, J., Danielsen, P.H., Esterhuizen, M., 662 Jacobsen, N.R., Vogel, U., Gant, T.W., Choi, I., FitzGerald, R., 2022. Reactive Oxygen Species in the 663 Adverse Outcome Pathway Framework: Toward Creation of Harmonized Consensus Key Events. Front. 664 Toxicol. 4, 887135. https://doi.org/10.3389/ftox.2022.887135 665 Tersago, K., De Coen, W., Scheirs, J., Vermeulen, K., Blust, R., Van Bockstaele, D., Verhagen, R., 2004. 666 Immunotoxicology in wood mice along a heavy metal pollution gradient. Environ. Pollut. 132, 385– 667 394. https://doi.org/10.1016/j.envpol.2004.05.029 668 Thévenod, F., Lee, W.-K., 2015. Live and Let Die: Roles of Autophagy in Cadmium Nephrotoxicity. Toxics 3, 669 130–151. https://doi.org/10.3390/toxics3020130 670 Tryphonas, H. (Ed.), 2005. Investigative immunotoxicology. Taylor & Francis, Boca Raton, FL. 671 Turley, A.E., Zagorski, J.W., Kennedy, R.C., Freeborn, R.A., Bursley, J.K., Edwards, J.R., Rockwell, C.E., 2019. 672 Chronic low-level cadmium exposure in rats affects cytokine production by activated T cells. Toxicol. 673 Res. 8, 227–237. https://doi.org/10.1039/c8tx00194d 674 24 Ulich, T., del Castillo, J., Guo, K., 1989. In vivo hematologic effects of recombinant interleukin-6 on 675 hematopoiesis and circulating numbers of RBCs and WBCs. Blood 73, 108–110. 676 https://doi.org/10.1182/blood.V73.1.108.108 677 Vallverdú-Coll, N., Mateo, R., Mougeot, F., Ortiz-Santaliestra, M.E., 2019. Immunotoxic effects of lead on birds. 678 Sci. Total Environ. 689, 505–515. https://doi.org/10.1016/j.scitotenv.2019.06.251 679 Van Brussel, K., Holmes, E.C., 2022. Zoonotic disease and virome diversity in bats. Curr. Opin. Virol. 52, 192– 680 202. https://doi.org/10.1016/j.coviro.2021.12.008 681 Villette, P., Afonso, E., Couval, G., Levret, A., Galan, M., Goydadin, A.-C., Cosson, J.-F., Giraudoux, P., 2020. 682 Spatio-temporal trends in richness and persistence of bacterial communities in decline-phase water 683 vole populations. Sci. Rep. 10, 9506. https://doi.org/10.1038/s41598-020-66107-5 684 Villette, P., Afonso, E., Couval, G., Levret, A., Galan, M., Tatard, C., Cosson, J.F., Giraudoux, P., 2017. 685 Consequences of organ choice in describing bacterial pathogen assemblages in a rodent population. 686 Epidemiol. Infect. 145, 3070–3075. https://doi.org/10.1017/S0950268817001893 687 Viney, M., Riley, E.M., 2017. The Immunology of Wild Rodents: Current Status and Future Prospects. Front. 688 Immunol. 8, 1481. https://doi.org/10.3389/fimmu.2017.01481 689 Vinken, M., Knapen, D., Vergauwen, L., Hengstler, J.G., Angrish, M., Whelan, M., 2017. Adverse outcome 690 pathways: a concise introduction for toxicologists. Arch. Toxicol. 91, 3697–3707. 691 https://doi.org/10.1007/s00204-017-2020-z 692 Wang, Z., Sun, Y., Yao, W., Ba, Q., Wang, H., 2021. Effects of Cadmium Exposure on the Immune System and 693 Immunoregulation. Front. Immunol. 12, 695484. https://doi.org/10.3389/fimmu.2021.695484 694 Wani, A.L., Ara, A., Usmani, J.A., 2015. Lead toxicity: a review. Interdiscip. Toxicol. 8, 55–64. 695 Watson, H., Videvall, E., Andersson, M.N., Isaksson, C., 2017. Transcriptome analysis of a wild bird reveals 696 physiological responses to the urban environment. Sci. Rep. 7, 44180. 697 https://doi.org/10.1038/srep44180 698 Wellcome Sanger Institute, 2023. Wellcome Sanger Inst. Blog. URL 699 https://sangerinstitute.blog/2023/09/14/1000-species-get-their-genomes-sequenced-for-the-first700 time/ (accessed 12.19.24). 701 World Health Organization, 2024. URL https://www.who.int/teams/environment-climate-change-and702 health/chemical-safety-and-health/health-impacts/chemicals (accessed 12.19.24). 703 Wu, Z., Lu, L., Du, J., Yang, L., Ren, X., Liu, B., Jiang, J., Yang, J., Dong, J., Sun, L., Zhu, Y., Li, Y., Zheng, D., Zhang, 704 C., Su, H., Zheng, Y., Zhou, H., Zhu, G., Li, H., Chmura, A., Yang, F., Daszak, P., Wang, J., Liu, Q., Jin, Q., 705 2018. Comparative analysis of rodent and small mammal viromes to better understand the wildlife 706 origin of emerging infectious diseases. Microbiome 6, 178. https://doi.org/10.1186/s40168-018707 0554-9 708 Zelikoff, J.T., Smialowicz, R., Bigazzi, P.E., Goyer, R.A., Lawrence, D.A., Maibach, H.I., Gardner, D., 1994. 709 Immunomodulation by Metals. Toxicol. Sci. 22, 1–7. https://doi.org/10.1093/toxsci/22.1.1 710 711 1 1. Introduction 1 Wildlife has been threatened for decades by several stressors from anthropogenic activities, which leads 2 to the sixth mass extinction (Birnie-Gauvin et al., 2017; Ceballos et al., 2017) and makes conservation of the 3 biodiversity one of the main issues of the century. Stressors are coming from many sources, including the 4 exposure to different kinds of pollutants such as non-essential trace metals (TMs) (for readability purpose, 5 this acronym and other used in the present paper are presented in Table 1). Contrarily to essential elements 6 that have important physiological functions in organisms, metals such as cadmium (Cd) or lead (Pb) are non7 essential elements to vertebrates since they are not known to have any biological function (Aras and Ataman, 8 2007). Transfer of these toxic metals from the environment to the organisms depends on several parameters 9 regulating their bioavailability (Lanno et al., 2004), defined as the proportion of the total concentration of a 10 chemical that is actually taken up from the environment (Naidu et al., 2008). Toxicological bioavailability refers 11 to the proportion of the concentration taken up that reaches toxicological targets (e.g., cells, tissues, organs) 12 through the circulatory system, subsequently causing a biological response (Lanno et al., 2004). Historically, 13 studies on TMs have mainly focused on accumulation in tissues and effects on classical ecotoxicological 14 endpoints (mortality, reproduction outputs, histopathology, biochemical, physiological disorders…) while 15 unconventional endpoints, such as impacts on the immune system or behaviour, have attracted more 16 attention only recently. Ecoimmunotoxicology refers to the understanding of the interactions between 17 exposure to pollutants, immune cells and the “delicate interactions associated with homeostasis mediated 18 by the immune system” (Tryphonas, 2005; Gallucci et al., 2020). 19 Table 1. 20 Exposure to non-essential TMs at toxic levels can affect the immune system through an up-regulation 21 (immunoactivation) or a down-regulation (immunodeficiency), both referring to immunomodulation (Zelikoff 22 et al., 1994; Devalloir et al., 2023). Immunodeficiency is defined as the decrease in effectiveness of the 23 immune response (Gallucci et al., 2020) while immunostimulation refers to the enhancement of the defensive 24 response, that can lead to auto-immune dysfunction or overall immune hypersensitivity (Luster and 25 Gerberick, 2010; McKee and Fontenot, 2016; Gallucci et al., 2020). These immunomodulation processes 26 Revised Manuscript with Track changes and Without Author Identifiers 2 mediated by toxic metals impact both innate and adaptive responses by unsettling the balanced mechanisms 27 of immune cell regulation (Tersago et al., 2004; Ebrahimi et al., 2020). Immunomodulation is also linked to a 28 myriad of external factors and stressors such as food shortage, life stage, or exposure to pathogens (Jackson 29 et al., 2009; Poulsen and Escher, 2012). Toxic effects also vary with species, sex, type (chronic or acute) and 30 route (ingestion, inhalation, dermal contact) of exposure (Gallucci et al., 2020). Immune cell dysfunction can 31 occur through ionic substitution of toxic metals due to similar affinity to receptors on organic compounds, 32 such as Cd2+ crossing Ca2+ channels (Choong et al., 2014). The main mechanism responsible for immune cell 33 damages is related to the toxicological impact of reactive oxygen species (ROS) including immune cell 34 membrane degradation (e.g., lipopolyperoxydation, Gera et al., 2015). Cadmium and Pb also have the ability 35 to upor downregulate inflammatory mediators and markers (cytokines, chemokines…), which can directly 36 modify the immune response (Ebrahimi et al., 2020; Devalloir et al., 2023). Cadmium and Pb can thus induce 37 differential effects on cytokine production through different signalling pathways (Thévenod and Lee, 2015). 38 For example, an increase of the pro-inflammatory response (TNF-α) has been shown in wood mice living near 39 a former smelter site contaminated by high Cd and Pb levels (Devalloir et al., 2023). 40 Since the emergence and the development of ecotoxicology in the 60’s, the discipline has evolved from a 41 relatively descriptive field to a much more mechanistic, hypothesis-based science. It aims at providing both a 42 better understanding of how pollutants can affect organisms at various spatial, temporal, and biological 43 organisation scales, and an improvement of risk assessment of chemicals released into the environment. 44 Among the approaches used in ecotoxicology to study the cascade (from molecular to population or 45 community levels) toxic effects of pollutants, adverse outcome pathways (AOPs) appeared as an innovative 46 and useful concept that could improve mechanistic approaches, here in ecoimmunotoxicology (Tryphonas, 47 2005; Baudiffier et al., 2024). Adverse outcome pathways represent a conceptual framework that highlights 48 existing knowledge linking a molecular initiating event (MIE) induced by a given toxic chemical to an adverse 49 outcome (AO) at multiple levels of biological organisation (Ankley et al., 2010). An AOP represents a 50 succession of key events (KEs) that can take various forms and depend on the information available. Molecular 51 initiating events are located at the molecular level and can be defined as “the initial points of chemical52 3 biological interaction within the organism” (Figure 1, Vinken et al., 2017). The relationship(s) between the 53 different events can be causal, mechanistic, inferential or correlation-based (Ankley et al., 2010). 54 Figure 1. 55 The present paper focusses on the effects of Cd and Pb on the immune system of mammals to propose 56 putative AOPs to test more mechanistic hypotheses in ecoimmunotoxicology of metals and to address future 57 questions in stress ecology (as other stressors than metals act on immunity) and ecotoxicology of free-ranging 58 mammals. Here, we focus on terrestrial mammals because reviews have recently been published on immune 59 toxicity of metals (or contaminants including metals) in fish (Lee et al., 2023, 2019), birds (Vallverdú-Coll et 60 al., 2019) and marine mammals (Desforges et al., 2016). All arsenic (As), Cd, Pb, and mercury (Hg) belong to 61 the 10 chemicals or groups of chemicals of major public health concern at a global scale (World Health 62 Organization, 2024). We however focus here on Cd and Pb only as they are more relevant than As or Hg to 63 terrestrial mammals in terms of exposure, and to keep the present article at a reasonable size. We based the 64 construction of our AOPs on reviewing the literature in the area, our own experience in ecoimmunotoxicology, 65 and different interactive databases related to AOPs. These databases provide open-source interfaces to 66 deposit AOPs, which facilitates the sharing of AOP knowledge and collaborative work (reviewed in Vinken et 67 al., 2017). 68 2. Materials and Methods 69 The articles presented in this paper have been selected according to the PRISMA statement (Moher et al., 70 2009). Literature search for study cases on terrestrial mammals was performed with Web of Sciences (June 71 2023) by using the following keywords, alone or in combination, full or truncated: immune system, immunity, 72 oxidative stress, AOPs/Adverse outcomes pathways, biomarkers, mammals (wild or captive), metals/heavy 73 metals, Pb/Lead, Cd/Cadmium, mixture effects. We also included other review articles about AOPs, 74 immunology and immunotoxicity in mammals. We ended up with 123 articles after the identification phase 75 and then selected the most pertinent articles based on their title and abstract (screening phase). After the 76 eligibility phase, we finally kept 51 articles, that were included in our synthesis. After reviewing the literature 77 and building pre-AOPs, we used four differentthe AOP-wiki interactive databases (https://aopwiki.org/, 78 10 exposure (Iavicoli et al., 2006; Wang et al., 2021). In the case of immunoglobulin production, studies showed 232 an inhibition of IgE and an inhibition of IgA and IgG in the case of Cd and Pb, respectively (Anetor et al., 2008; 233 Wang et al., 2021). 234 Figure 4. 235 4.6. Evidence of the effects of a Cd-Pb mixture on the immune system of mammals 236 based on an AOP framework Development of an AOP for the effects of a Cd-Pb mixture on 237 the immune system of mammals 238 In the environment, organisms are generally exposed to a mixture of contaminants rather than to a single 239 trace metal. These multiple exposures can be detrimental to wild vertebrates even if contaminants are 240 present at low concentrations. This concept is called “cocktail effect” or “mixture toxicity” (Beyer et al., 2014). 241 Ecotoxicological studies have mainly been performed through a chemical-by-chemical approach of exposure, 242 while wildlife are exposed to “dynamic chemical mixtures” through their life (Morrissey et al., 2023). It exists 243 only a few papers that relate mixture effects of trace metal elements (Cd, Pb) on the immune system. Overall, 244 Cd and Pb produce similar effects in vertebrates with an outcome resulting in either an immunostimulation 245 or an immunodeficiency. HoweverThus, little is known about their additive, synergetic and/or antagonistic 246 effects on the immune system of vertebrates. Synergetic and antagonistic effects are the results of the 247 competition among metals for organic and inorganic ligands during accumulation and uptake (Altenburger et 248 al., 2003; Belden et al., 2007; Balistrieri and Mebane, 2014). It has been shown that Cd-Pb mixture can 249 enhance the production of white blood cells in mice, indicating the activation of the immune system in 250 response to the exposure (Cobbina et al., 2015). In addition, these authors found that the exposure to a 251 mixture result in a global antagonistic effect, in comparison to their individual toxic effect. One hypothesis 252 could be that Cd and Pb both produce thiols such as GSH, enabling a protection of the cell from their toxicity. 253 However, other antioxidant markers (superoxide dismutase, glutathione peroxidase) were lower when mice 254 were exposed to a Cd and Pb mixture, potentially creating higher oxidative stress in immune cells. Another 255 study in male rat showed that the exposure to a mixture of eight trace metals, including Cd and Pb, induces 256 a suppression of both humoral and cell-mediated immune response (Jadhav et al., 2007). Fortier et al. (2008) 257 11 depicted a non-significant effect of the Cd-Pb mixture in human leucocytes (viability and apoptosis 258 proportion). These few studies provide different conclusions, suggesting that a variety of pathways could be 259 involved in the response to Cd-Pb mixture. However, we still lack information about how cytokines signalling, 260 TH1/TH2 polarization of B-cells and antibody production could be modulated by this mixture. Here, we propose 261 a potential AOP for Cd-Pb mixture on the immune system of mammals with potential additive effects or 262 synergetic and antagonistic interactions (Figure 54), based on the few studies presented earlier as well as on 263 hypotheses and deductions from our single-chemical AOPs (Figures 2, and 3 and 4). 264 265 12 Figure 54. 266 5.7. Gaps in current knowledge and research perspectives 267 In the present article, we built putative AOPs depicting the cascade effects of Cd and Pb alone, and in 268 combination, on the immune system of mammals. We identified several pathways that could lead to immune 269 dysfunctions in mammals. However, despite identifying links between contaminants, MIEs, KEs and AOs using 270 the existing knowledge, how the different cascades of reactions occur in wild organisms, and what adverse 271 outcomes arise at various levels of biological organisation remain uncertain in wildlife since most of the work 272 has been done in lab animals. To fill this gap, we identified several biomarkers indicated in bright pink boxes 273 (and indicated by an asterisk) in Figures 2 to 6, and develop hereinafter the research priority identified from 274 these putative AOPs, the literature and our own experience in immunotoxicity of metals in free-ranging small 275 mammals (Devalloir et al., 2023; Powolny et al., 2023; ongoing research). 276 One of the primary aspects to consider, common to both Cd and Pb exposure, relies upon their impacts 277 through NF-κB. This family of inducible transcription factors regulates multiple aspects of innate and adaptive 278 immune functions and is a pivotal mediator of inflammatory responses (Liu et al., 2017). NF-κB induces 279 several functions in innate immune cells, including maturation of dendritic cells, activation and differentiation 280 of inflammatory T cells, polarization of macrophages and their production of pro-inflammatory 281 cytokines/chemokines, and recruitment of and anti-apoptotic effect on neutrophils. These cells express 282 pattern-recognition receptors that detect various microbial components (so-called pathogen-associated 283 molecules patterns, PAMPs) as well as molecules released by necrotic cells and damaged tissues (co-called 284 damage-associated molecular patterns, DAMPs). Thus, a disruption of NF-κB by contaminants (here Cd or Pb) 285 could alter these functions and might lead to a lower ability of the innate immune system to fight against 286 pathogens. NF-κB also exerts a central role in mediating T cell receptors signalling and naive T-cell activation, 287 which is an important component of the adaptive immune system. Deregulation of NF-κB activation can cause 288 aberrant T-cell activation, which might lead to associated autoimmune and inflammatory responses. 289 Inflammation is a protective response of the host to infections and tissue damages and is normally beneficial 290 and resolved in a timely manner. Given that wild rodents have been shown to be particularly exposed to 291 infections, deregulation of NF-κB activation by metals might be particularly harmful to wild animals, both by 292 13 lowering their ability to control pathogenic infections and through excessive or long-lasting tissue damages, 293 ultimately leading to acute or chronic inflammatory diseases (Liu et al., 2017). Studying this pathway, 294 especially in free-ranging wildlife, is however difficult due to the complexity of the activation of NF-κB, which 295 involves two major pathways (canonical and non-canonical) related to different signalling mechanisms. The 296 first pathway responds to diverse stimuli including ligands of various cytokine receptors, pattern-recognition 297 receptors (PRRs), TNF receptor (TNFR) superfamily, B and T cell receptors. In contrast, the non-canonical 298 pathway selectively responds to a specific group of stimuli, including ligands of a subset of TNFR superfamily 299 members (Liu et al., 2017). Furthermore, the activity of NF-κB depends on three layers of regulation. A first 300 layer refers to the dynamic nature of the NF-κB system, which is never static and is schematically under 4 301 forms of activity (OFF, lowand high-ON, and constitutively active). In “OFF” state, the NF-κB activity is usually 302 too low to be reliably detected (Meier-Soelch et al., 2021). Constitutively active state refers to significant NF303 κB activity found in specific cells such as Sertoli cells, B cells, hair follicle cells and neurons, making it difficult 304 to be studied in wildlife. “Low ON” state refers to a continuous low-grade activation that occurs in situations 305 of smoldering inflammation or in tumour microenvironment, and “high ON” state is a fast and transient 306 activation in response to proinflammatory triggers. Another layer of regulation is related to the gene specific 307 recruitment of NF-κB and a third one involves a remarkable cell to cell variability, with cells having low NF-κB 308 activity while adjacent cells exhibit high activity and strong activation of NF-κB target genes (Meier-Soelch et 309 al., 2021). These authors therefore warn that (cell) population-based assays can be potentially misleading. 310 The NF-κB activation rarely occurs in isolation and frequently other stress pathways such as p38 mitogen311 activated protein kinases (p38 MAPKs), c-Jun N-terminal kinases (JNKs) or JAK-STAT are activated in parallel. 312 As all have been shown to be involved in metal-related toxicity in human or lab-animal models (but not in 313 AOP framework to the best of our knowledge), they are also good candidate models to be further studied in 314 metal immunotoxicology. 315 Studying the responses of the exposure to Cd and/or Pb on the innate immune system involving NF-κB 316 pathways would rely upon (i) an estimate of the exposure to the above-mentioned metals, (ii) the potential 317 related increase of ROS production, (iii) the activation of the NF-κB pathway, (iv) the increase of some 318 cytokines, (v) the modulation of more general mechanisms (increase of haematopoiesis, increase of 319 14 neutrophils, decrease of phagocytosis activity…), and (vi) ultimately adverse outcomes such as the increase 320 of infection rate. Estimating the exposure of wildlife to contaminants in all its complexity (contaminant 321 mixtures, spatio-temporal variations, speciesor stage-specificity, ecological context) is still a hard task 322 recently discussed (Morrissey et al., 2023). However, quantifying the exposure to metals and metalloids using 323 e.g. ICP methods upon invasive (internal organs such as liver or kidneys) or less invasive (blood, hair…) of a 324 species of interest along a pollution gradient is relatively efficient. Similarly, measuring ROS production (free 325 radicals such as superoxide anion or nitric oxide, non-radical ROS such as hydrogen peroxide), oxidative 326 damage (dROMS, TBARS, protein carbonyls) and enzymatic (SOD, GPx, catalase, GST, GR) and non-enzymatic 327 (OXY, GSH) anti-oxidant defences is now allowed by a number of assays that can be relatively easily performed 328 on tissues or less invasive fluids like blood (Costantini, 2022; Tanabe et al., 2022). The study of the NF-κB 329 pathway and its three layers of regulation (activity, gene specific recruitment, cell to cell variability) is far more 330 complex in wildlife. Meier-Soelch et al. recently reviewed the concepts and the methods related to the 331 monitoring of levels of cellular NF-κB activation states (Meier-Soelch et al., 2021). Most of these techniques 332 require high biochemical skills and appropriate lab equipment, and furthermore need cell isolation and/or 333 cell culture to take into account the high variability of NF-κB activation state among cell types and induction 334 kinetics. Most of this might be out of reach to wildlife toxicologists, at least on a short term. The relatively 335 recent development of transcriptomic approaches in ecology and ecotoxicology, however, could bring 336 important insights into the involvement and the regulation of NF-κB pathways in wildlife exposed to 337 pollutants. An example, if not of a true ecotoxicological study but at least of a comparison of environments 338 differing by pollution level (among other parameters), is a recent analysis of the transcriptome of rural and 339 urban great tit (Parus major) populations in both blood and the liver. The study revealed that the NFKBID 340 gene (coding for the protein called nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, 341 delta; IκBNS) was one of the 20 most significant annotated genes that were differentially expressed between 342 urban and rural birds in whole blood transcriptomes (Watson et al., 2017). Similar RNA-sequencing analyses 343 could be performed in wildlife populations inhabiting habitats with contrasted levels of Cd and/or Pb, or in 344 individuals exposed under semi-natural controlled conditions in mesocosms. Complementarily to the random 345 search of differential expression of certain genes among exposed and nonor less-exposed populations or 346 15 individuals from the global transcriptomes, the present AOPs allow to specifically look for possible differential 347 expression of genes related to NF-κB pathways. Such approach of course requires that the genes and RNAs of 348 interest have been sequenced and identified in databases, but the number of fully sequenced genomes 349 increases fast. The full genome of 1,000 species is available within the Tree of Life Programme (Wellcome 350 Sanger Institute, 2023) and overall it is estimated that the genome of more than 3,000 species has been fully 351 sequenced, with however large disparities in genomic representation (e.g. overrepresentation of vertebrates 352 over arthropods) and assembly quality (gene annotations available for 34% of the taxa, Hotaling et al., 2021). 353 To go further into the key events of the present AOPs, the potential increase of some cytokines under Cd 354 and/or Pb exposure would be required but the lack of species-specific monoclonal antibodies has historically 355 been a barrier to integrating free-ranging mammals, beyond the well-known lab model species such as lab 356 mice (Mus musculus) and rats (Rattus norvegicus). Even if the availability of reagents developed in lab mice 357 has interestingly been used to study the immunity of their wild counterparts (Abolins et al., 2018, 2017, 2011), 358 the immune system of most of wild species cannot be studied with that resolution because monoclonal 359 antibodies developed in lab mice do not (or hardly) cross-react with other rodent species, even 360 phylogenetically closed (e.g. Apodemus sylvaticus; Devalloir, 2023). Some cytokines and other immune 361 molecules such as immunoglobulins can however be quantified through the use of e.g. Elisa kits that 362 sometimes give satisfactory results among species (Devalloir et al., 2023). However, here again, the 363 development of full genome and transcriptome sequencing is a promising perspective as it offers the 364 opportunity to study genes or their expression rather than the proteins themselves, or produce recombinant 365 proteins for protein-based immunology in species of interest (Flies et al., 2020). Flies et al. argue for “a 366 systematic effort to develop and characterize antibodies, nanobodies, or aptamers that bind to conserved 367 protein motifs across taxonomic orders” to overcome the funding constraints limiting the development of 368 specific reagents for all species of interest (Flies et al., 2020). The study of other key events described in the 369 present AOPs such as the increase of neutrophils or the decrease of phagocytosis activity can be performed 370 by relatively classical -but usually fluidand time-consumingtechniques such as blood smears and 371 phagocytosis assays, microbial killing assays (MKA) or whole blood microbial killing assay (WBMKA), 372 respectively (Devalloir 2023). 373 16 Regarding the adaptive components of the immune system, exposure to Cd or Pb has been shown to 374 decrease CD4+/CD8+ ratio and MHCII molecules expression, affecting the activation of TH1 and TH2 cells and 375 reducing humoral responses. T lymphocytes (T cells) are divided into two major cell types: T helper (TH) and 376 T cytotoxic (TC) cells. They can be distinguished by the presence of CD4 (for TH) and CD8 (for TC) membrane 377 glycoproteins on their surfaces. The ratio of CD4+ to CD8+ T cells is approximately 2:1 in peripheral blood of 378 healthy human and mouse, while a decreased ratio is usually an indication of immunodeficiency or 379 autoimmune disease, aging or inflammation. T cells express a unique antigen-binding receptor (T-cell 380 receptor, TCR), which recognizes processed pieces of antigen (usually peptides) bound to cell membrane 381 proteins, the MHC. Naïve CD4+ TH cells, when they bind to an MHCII-peptide complex, become activated, 382 proliferate and differentiate into various effector T-cell subsets, mainly T helper cells (type 1 (TH1) and type 383 17 (TH17)), and T helper type 2 (TH2) and T follicular (TFH) cells. TH1 and 17 regulate responses to intracellular 384 pathogens while TH2 and TFH regulate responses to extracellular pathogens such as bacteria and parasitic 385 worms. Naïve CD8+ TC cells, when they bind to an MHCI-peptide complex, become activated, proliferate and 386 differentiate into cytotoxic T lymphocyte (CTL), which have a vital function in recognizing and eliminating cells 387 displaying non-self antigen-MHCI complex (such as virus infected or tumour cells). Further research should 388 thus also focus on cell populations, and more precisely the CD4+/CD8+ ratio, and associated major 389 histocompatibility complex (MHC) molecules (Punt, 2019). Such immunophenotyping can be done through 390 flow cytometry but this would require species-specific antibodies as anti-mouse antibodies do not or hardly 391 cross-react with immune cells from other rodent species (Devalloir, 2023; García-Mendoza et al., 2021). As 392 for some of the perspectives mentioned above, the development of antibodies that bind to conserved protein 393 motifs across taxonomic orders, if not the development of specific reagents for some species of interest, is 394 required, as we do not see how the CD4+/CD8+ ratio could be measured otherwise. Complementarily, a 395 transcriptomic approach can also be developed as it has recently been done in sheep inoculated with a 396 virulent Anaplasma phagocytophilum to study the temporal patterns of gene expression in response to the 397 inoculation (Eskeland et al., 2023). In the latter, however, the transcriptomic approach was completed by 398 flow‑cytometry of peripheral blood mononuclear cells to further study the changes of CD4+/CD8+ ratio. 399 17 The putative AOPs proposed in the present articles focus mainly on sub-individual levels of organisation, 400 and perspectives of research aiming at better understanding immunotoxicological mechanisms at these levels 401 of organisation have been presented above. The AOP approach, in ecotoxicology and by extension here in 402 ecoimmunotoxicology, aims to also address effects of pollutants at higher levels of organisation. A deficient 403 immune system at the sub-individual level could lead to severe impacts on individual fitness and population 404 dynamics. The hypothesis that reduced immune function and/or increased oxidative stress in wildlife 405 chronically exposed to pollutants might increase parasite prevalence and/or loads and impact fitness has 406 poorly been investigated. It has been shown that wild rodents exhibited significant differences in the 407 functioning of their immune system compared to lab animals. (Viney and Riley, 2017) recently reviewed this 408 issue and stated that wild rodents compared to laboratory animals exhibit (i) much higher humoral (antibody) 409 responses, (ii) extensive antigenic exposure as revealed by cellular immune system, (iii) depressed 410 proliferative and cytokine responses in ex vivo-stimulated immune cells. Studying exposure of wildlife to 411 pathogens is increasingly accessible through e.g. high-throughput DNA sequencing, allowing for the 412 identification of a number of zoonotic or non-zoonotic, emerging or well-known, external or internal 413 pathogens and parasites (e.g., Diagne et al., 2016; Villette et al., 2020). For assessing bacterial communities, 414 a mix of organs allows to get a better view of bacterial assemblages that single organs (Villette et al., 2017) 415 but this unfortunately requires the sacrifice of (some) individuals, which might be ethically questionable when 416 working on protected and/or endangered wildlife. The use of fluids like blood or faeces is less invasive but 417 would give a partial view of the bacterial assemblages or even false negatives as their presence in the fluids 418 depends on their distribution between organs and fluids, or on their excretion kinetics. Viruses also are 419 increasingly studied in free-ranging vertebrates (Van Brussel and Holmes, 2022; Wu et al., 2018) and nonor 420 little-invasive techniques based on faecal and oropharyngeal swabs have recently been developed (Bergner 421 et al., 2019). 422 Ecotoxicological studies often face a lack of environmental and ecological representativeness regarding 423 the overall exposure of wildlife, and how pollutants are reacting in the environmental matrix. Morrissey et al. 424 (2023) recently address this question to improve wildlife risk assessment. Indeed, the AOP framework 425 answers some of the specific guidelines they proposed (Morrissey et al., 2023). First, building AOPs allows a 426 18 consideration of the general chemical complexity (i.e., mixture effects, chemical properties, metabolites) of 427 pollutants. Each compound has its own mode(s) of action and pathways that lead to one or several toxic 428 effects at different levels of biological organisation. Identifying molecular initiating event(s) and key event(s) 429 that lead to adverse outcome(s) participates to understand how complex are the interactions between 430 pollutants, here Cd and Pb, in organisms. Second, once AOPs are built and the main pathways identified, it is 431 possible to make it speciesor life-stages specific (i.e., considering organismal complexity; Morrissey et al., 432 2023). This will guide the formulation of hypothesis and the choice of biomarkers. This (new) framework 433 participates in answering new questions and guide researchers into new directions in ecotoxicology, stress 434 ecology and ecoimmunotoxicology. Building AOPs before starting an experiment appears as something that 435 should become more common in these field of research to better apprehend mechanisms underlying toxic 436 effects and identify the research gaps (Baudiffier et al., 2024). In addition, developing the knowledge of the 437 database presented by Vinken et al., (2017) would encourage data sharing and literature findings to better 438 direct further research. To date, the most important limitations in making AOPs about a specific subject is the 439 lack of literature that can restrict the whole process. 440 441 6.8. Conclusion 442 Adverse outcome pathways are an efficient method to identify the mechanisms underlying the effects of 443 pollutants at various levels of biological organisation and thus, to identify research gaps in ecotoxicology and 444 risk assessment. Here, we identified potential MIEs and KEs that lead to AOs related to immune disorders in 445 mammals in response to Cd and/or Pb exposure. Cadmium and Pb may induce immuno-stimulation and 446 immunodeficiency through different pathways of the immune system, impacting both innate and adaptive 447 immune responses. Our findings could guide future research in ecoimmunotoxicology of metals. There are 448 still large open questions about how organisms, at different levels of organisation, react to pathogens and 449 infections in a context where exposure to pollutants affects immunity. Mixture effects of trace metal are also 450 still largely understudied. It is now of major importance to understand how these pollutants interact with 451 wildlife to improve conservation policies. We identified potential synergetic and antagonistic interactions 452 19 produced by the co-exposure to Cd and Pb, which also brings questions about the potential adaptation of 453 wild populations to those mixtures, in comparison to single-pollutant exposure. 454 Acknowledgements 455 The authors thank Michaël Coeurdassier, Clémentine Fritsch, and Francis Raoul, for fruitful discussion about 456 the present article. 457 Author contributions 458 Conceptualization: CH, RS; Data curation: CH; Formal analysis: CH; Funding acquisition: RS; Investigation: CH; 459 Methodology: CH; Project administration: RS; Supervision: QD, NvdB, RS; Validation: CH, QD, CG, NvdB, RS; 460 Visualization: CH; Writing - original draft: CH, RS; Writing - review & editing: CH, QD, CG, NvdB, RS. 461 Declaration of competing interest 462 The authors declare no competing interests. 463 Funding information 464 This study is funded by the CHRONIC project (No. 956009) ‘Chronic exposure scenarios driving environmental 465 risks of Chemicals’, an Innovative Training Network (ITN) funded by the European Union's Horizon 2020 466 research and innovation program under the Marie Skłodowska-Curie Actions (MSCA). 467 468 2 Table 2. Key-events found in the AOP-wiki database in relation to the AOPs developed in the present article. 4 Individual Event ID are indicated in each associated box of the AOPs (see Figures). The last column contains 5 the ID of the other AOPs from the AOP-wiki database that include the different key events of our own AOPs. 6 Event ID KE name Biological level Other AOPs 1238 Activation of oxidative stress pathway Molecular 1969 Increase, Oxidative Stress Molecular 437/457/459/507/509/510/511 1770 Decrease, Mitochondrial membrane potential Cellular 328/387/447 1262 Apoptosis Cellular 205/207/212/285/419/439/452/ 393/476/460/500/491/502 1365 Increase, Apoptosis Cellular 322/331/330/325/326/441/444/509 1172 Increased activation, Nuclear factor kappa B (NF-κB) Cellular 382/377/319/443 130 Depletion, GSH Cellular 492 151 Activation, Inflammatory cytokines, chemokines, cytoprotective gene pathways Molecular 1020 Suppression, IL-2 and IL-4 production Cellular 154 1225 Immune system inflammation Tissue 1750 Increased inflammatory immune responses Tissue 320/426 149 Increase, Inflammation Cellular/Tissue 27/115/206/280/439/505 1569 Impaired T cell activation Cellular 1702 Suppression of T cell activation Cellular 227 1644 Impaired Ab production Cellular 403 Suppression, Immune system Individual 14/84/85/432 7 1 Figure 1. Generic AOPs structure for a single chemical, that can be applied to any organism and spread for 1 different biological organisations (adapted from Vinken et al., 2017). MIE stands for molecular initiating 2 events, KE for key events and AO for adverse outcome. 3 4 5 6 Figure 2 Figure 2. Representation of an adverse outcome pathway for cadmium toxicity on the mammal immune 7 system. As in Figure 1, the chemical initiator is represented in orange, molecular initiating events in green, 8 key events in pink and adverse outcomes in red. Bright pink (also indicated with a black star) key events are 9 potential biomarkers that could be used in future applied studies. The M and N abbreviations above some 10 boxes stand for macrophages and neutrophils, respectively. Numerical Event ID next to the different 11 components of the AOP are from https://aopwiki.org/, and are also presented in Table 1. 12 *The modulation of TNF-α vary depending on cadmium exposure time. Acute exposure result in an increase 13 of its production while a longer exposure time resulted in a decrease of TNF-α. 14 15 16 17 18 19 20 21 22 3 Figure 3. Representation of an adverse outcome pathway for lead toxicity on the mammal immune system. 23 The chemical initiator is represented in orange, molecular initiating events in green, key events in pink and 24 adverse outcomes in red. Bright pink (also indicated with a black star) key events are potential biomarkers 25 that could be used in future applied studies. The M and N abbreviations above some boxes stand for 26 macrophages and neutrophils, respectively. The dotted arrows indicate that the link between two events is 27 not precisely detailed in the literature. Numerical Event ID next to the different components of the AOP are 28 from https://aopwiki.org/, and are also presented in Table 1. 29 30 31 32 4 Figure 4. Representation of an adverse outcome pathway for both lead and cadmium toxicity on the mammal 33 immune system. The molecular initiating events are represented in green, key events in pink and adverse 34 outcomes in red. Bright pink (also indicated with a black star) key events are potential biomarkers that could 35 be used in future applied studies. The dotted arrows indicate that the link between two events is not precisely 36 detailed in the literature. Numerical Event ID next to the different components of the AOP are from 37 https://aopwiki.org/, and are also presented in Table 1. 38 39 40 5 Figure 5. Representation of an adverse outcome pathway of the mixture of cadmium and lead toxicity on the 41 mammal immune system. The chemical initiator is represented in orange, molecular initiating events in green, 42 key events in pink and adverse outcomes in red. Bright pink key events (also indicated with a black star) are 43 potential biomarkers that could be used in future applied studies. (=) are for additive interaction, (+) are for 44 synergetic interaction, and (-) are for antagonistic interaction. Numerical Event ID next to the different 45 components of the AOP are from https://aopwiki.org/, and are also presented in Table 1. 46 47 48 Declaration of interests ☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. ☐ The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Declaration of Interest Statement